Avaya CPSEE_TSP500 - Manual (Page 49)

Chapter 4 Features and Enhancements
CPSEE_TSP500 User Guide Version 4.0 Page 4-7
This document is confidential and proprietary to SER Solutions and is not for external use.
Text-To-Speech (TTS) Messages - Feature
The ability to play an arbitrary TTS voice message has been added to the
TSP500. The creation of the TTS message is not the responsibility of the
TSP500, but the TSP500 must have read access to the message. Access to the
message is provided by a Samba mount to an external drive. The external
drive is mounted locally to the path /home/tts. To ensure the drive is
mounted, an entry in the /etc/fstab file must be inserted. The format of the
entry is:
/// directory path> username=,password=word>, 0 0
Example:
//10.2.109.4/tts /home/tts smbfs user-
name=eisadmin,password=admin,defaults 0 0
The term “defaults” equivocates to “rw,suid,dev,exec,auto,nouser,async” per-
missions.
The Adjunct needs to know the audio output format in order to create a
voice message that can be played. The current Protocol C Message to send
resource information to the Adjunct has been modified to include encoding
format, bandwidth, and bit resolution. When the Adjunct starts and requests
resource information, an additional resource message
(M_GET_RESOURCES) with a mode value of 4 is sent. This message is
sent prior to the resource message with a mode value of 3 for backward com-
patibility. This message includes three data fields for TTS voice specific
parameters. The first “data” field contains the encoding format and can have a
value of 0 for Mu-Law and 1 for A-Law. The second “data” field contains the
bandwidth and is always 8000, since a bandwidth of 8K is only supported.
The third “data” field contains the bit resolution and is always 8, since a bit
resolution of 8-bit is only supported.
When play voice message request (M_MSG_PLAY) is received, the TSP500
determines if the request can be performed. The “flag” field descriptor with
the number one indicates the voice message is to be played immediately. The
play immediate option indicates the message is to be played immediately
using the trunk’s DSP. Otherwise, the message is placed on a one-second-
timer queue and an Encore DSP is used to play the message. The “key” field
descriptor may contain a voice message name or (optionally, a relative path
and) a filename and extension. If the “key” has an extension, a search of the
Text-To-Speech directory (/home/tts) for the existence of the “key” is per-
formed. Otherwise, a search of the Encore directory (/usr/vox) for the “key”
plus “wav” extension is performed.
When the non-interruptible voice message completes playing, a route request
(M_ROUTE_REQ) is sent to the Adjunct with a result code indicating the
voice message completed (SWR_MSG_PLAY_DONE).